🚨 BREAKING: Google Announces ChatGPT Competitor (12 Highlights)
Quick summary and thoughts on Google's MAJOR announcement
No sooner was SmokeBot done firing off the send today and DOMINATING a burrito bowl from Chipotle… and Google goes ahead and announces BARD— ostensibly its ChatGPT competitor.
As a rule, we don’t want to flood your inbox, but this is landscape-changing news and we didn’t want to wait until Wednesday.
Here’s what you need to know.
Google CEO Sundar Pichai announces Bard, an AI powered search tool, in a blog post
Here’s the nut:
Google will provide more details at their “Google Presents” event on Wednesday. We’ll have it covered for you.
Here is a link to the full blog post from today, but we are going to break it down in 12 highlights so you can digest this thing.
1) “we re-oriented the company around AI six years ago”
Right off the bat, “Pich-AI” lets it be known that Google is now an AI company.
Partially true? Yes, of course.
Would that phrase be coming out of his mouth at this point if not for the release and success of ChatGPT? No.
2) their mission: “organize the world’s information and make it universally accessible and useful”
There’s a book called The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail.
SmokeBot is certainly not here to say that Google is going to fail, but the re-stating of the mission makes it clear that they view AI (and Bard) as a way to improve, supplement, and perhaps protect their search business. This is why the features you’re about to read about are all search-focused.
But what if the AI revolution isn’t just about “organizing” and making information “accessible”, but rather about “creating”?
Something to think about.
3) “the scale of the largest AI computations is doubling every six months, far outpacing Moore’s Law”
Moore’s Law says that computing power doubles every two years. Google says that speed is actually 6 months with AI.
Imagine, then, how quickly things will improve if the capabilities we see today DOUBLE by summer in the Northern Hemisphere.
4) “fresh, high-quality responses… learn more about the best strikers in football right now”
A clear dig at ChatGPT, which is trained on data through 2021 and still serves Her Majesty, The Queen of England… for now.
As we noted in the send earlier, Microsoft’s New Bing may debut with the newest version of ChatGPT by Wednesday. And it will presumably include up-to-date results. So this may be a very short-lived advantage.
Not even Beta. Not Alpha. Experimental. This is a shield for when it inevitably gets something grotesquely wrong. Google has more reputational risk than OpenAI and Bing 😭.
6) “lightweight model version of LaMDA… this much smaller model requires significantly less computing power, enabling us to scale to more users, allowing for more feedback”
In short, they are not releasing the full thing. So this means one of two things:
1) They have preached caution and don’t want to release their most advanced tech until the world is ready for it.
2) It’s a hedge. So if Bard sucks, they can say they have something better.
7) “meet a high bar for quality, safety and groundedness in real-world information”
I’d argue this is another dig at OpenAI’s more… liberal approach to releasing AI. But, like Apple and privacy, Google seems to be taking the adult in the room approach with AI.
8) “we’re working to bring [language, image, and music] AI advancements into our products, starting with Search”
As we’ve noted before, Google is working on image, video, and music generation AI.
9) “safe and scaleable” APIs for developers
While ChatGPT gets all the pub, it’s OpenAI’s APIs, which allow developers to build apps atop their technology, that may be the real game-changer.
Google is making it clear they will play that game, too, but do so in a more measured way.
10) “bring experiences rooted in these models to the world in a bold and responsible way”
OK now they let the PR guy have too much fun.
When was the last time you ever met someone who is Bold and Responsible?
Tom Cruise jumping out of an airplane 80 times to get the next scene right is bold, but it’s not responsible.
Going to bed at 10PM is responsible, but it’s hardly bold. Bold is partying until 2AM, watching a few episodes of Family Guy, eating a bag of popcorn, and downing two hard seltzers, all to wake up at 6:12AM to get started on the latest SR newsletter. THAT’S bold.
Anyway, you get the point. Hard to be both, Google.
11) “turning to us for quick factual answers, like how many keys does a piano have?… but increasingly, people are turning to Google for deeper insights and understanding”
Basically, Google doesn’t want to provide just facts. It wants to provide detailed, nuanced answers to queries, with context, in a natural-language format.
The question, as it is with ChatGPT, is where does the information come from?
If you thought creators and publishers were bent out of shape over ChatGPT and image apps, like Stable Diffusion and MidJourney, “training” on their data and remixing it without credit, how will website owners, who rely on Google for views, react when Google remixes the content atop search results?
[They already do this with snippets, but Bard sounds like snippets on steroids.]
12) “soon, you’ll see AI-powered features in Search that distill complex information and multiple perspectives into easy-to-digest formats”
Yep, snippets on steroids sounds about right.